INDEX

Explanations

elements related to recognition and the impact of educators

oai_token-act-pair · gpt-4o-mini Triggered by @bot

concepts and foreign words

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RegressionTest

-0.79

irvana

-0.61

principalColumn

-0.56

principalTable

-0.56

fillType

-0.55

delwed

-0.55

webElementXpaths

-0.53

tagHelperRunner

-0.51

 ModelExpression

-0.50

Bef

-0.48

POSITIVE LOGITS

 localidad

0.39

 Economía

0.38

 églises

0.38

 elettrica

0.37

Lycka

0.37

droje

0.35

 risorse

0.34

 desierto

0.34

 documentación

0.33

ſelf

0.33

Activations Density 0.017%

elements related to recognition and the impact of educators

concepts and foreign words

No Comments

No Known Activations